Submacular Hemorrhage in 2026: Anti-VEGF, Pneumatic Displacement, rtPA, or Vitrectomy?

Research Snapshot: This 2026 narrative review synthesizes evidence on submacular hemorrhage (SMH), primarily secondary to neovascular age-related macular degeneration (nAMD) and polypoidal choroidal vasculopathy (PCV). Anti-VEGF therapy remains essential for controlling the underlying neovascular disease and is a reasonable less-invasive strategy for small-to-moderate hemorrhages. For larger, thicker, or visually threatening hemorrhages, pneumatic displacement (PD), often with recombinant tissue plasminogen activator (rtPA), may accelerate foveal blood clearance. Pars plana vitrectomy (PPV) with subretinal rtPA can achieve substantial anatomical displacement in selected complex cases, but randomized comparisons have not demonstrated consistently better vision than less-invasive displacement. In the 90-eye STAR trial, six-month gains were 17.2 ETDRS letters after PPV versus 15.4 after PD (P = .776). The review emphasizes early individualized intervention, careful complication counseling, and the absence of a validated universal treatment algorithm.

Article type: Narrative review of recent clinical studies, randomized trials, observational cohorts, and previously published meta-analyses; it is not a new patient cohort, systematic review, or original pooled analysis. The authors primarily address SMH associated with nAMD and PCV, rather than traumatic, myopic, or retinal-artery-macroaneurysm hemorrhage.

SMH places blood between the neurosensory retina and retinal pigment epithelium (RPE). Proposed damage mechanisms include separation of photoreceptors from metabolic support, iron-mediated oxidative toxicity, inflammation, and mechanical effects of the clot. These mechanisms motivate prompt assessment, although the optimal time window and intervention remain uncertain. The Fight Retinal Blindness! registry incidence cited in the review is 4.6 SMH events per 1,000 treated nAMD eyes annually; this is a registry-derived estimate, not a finding generated by the present review.

Intravitreal anti-VEGF addresses active neovascularization and remains central whether or not blood displacement is performed. The review cites a 2025 meta-analysis of 43 observational studies (1,415 eyes in the summarized comparison) reporting no statistically significant difference in pooled visual improvement between anti-VEGF monotherapy and surgical approaches (P = .19), with reported retinal detachment rates of 0.1% versus 10.6%, respectively. These complication percentages come from nonrandomized, clinically heterogeneous treatment groups and should not be interpreted as causal relative risks.

For hemorrhages larger than approximately three disc diameters, evidence is mixed. One 31-eye retrospective comparison found no significant 12-month BCVA difference between monotherapy and PPV; another 66-eye multicenter cohort found greater adjusted visual gain with surgery despite worse baseline hemorrhage, although final acuity did not differ significantly. There is no established anti-VEGF agent proven superior specifically for SMH.

PD uses an intravitreal expansile gas bubble, commonly SF6 or C3F8, to move subretinal blood away from the fovea; anti-VEGF is commonly continued. In a 47-eye retrospective PCV study with large SMH, aflibercept plus PD yielded a gain of at least three lines in 60.0% versus 18.2% with aflibercept alone at 12 months. However, another retrospective comparison found faster blood displacement without a statistically significant difference in final 12-month visual acuity.

The benefit may be greatest for selected larger or thicker, recent-onset hemorrhages, but studies use different entry criteria, hemorrhage definitions, gas regimens, and outcome measures. Faster anatomical clearance should not be equated automatically with superior final vision.

rtPA is used to facilitate clot liquefaction and can be delivered intravitreally alongside PD or subretinally during PPV. In the TAPAS randomized factorial trial (56 eyes), adjunctive intravitreal rtPA was associated with approximately 16 additional ETDRS letters at three months versus treatment without rtPA (P = .02), while C3F8 gas alone did not demonstrate the same visual benefit. The review reports more rapid hemorrhage clearance with rtPA.

In a 127-eye retrospective study of PD plus anti-VEGF with or without rtPA, complete displacement rates were 92.1% and 90.9%, with comparable visual gains; breakthrough vitreous hemorrhage was more common without rtPA (reported OR 7.03; P = .004), particularly in larger hemorrhages. Other meta-analyses summarized by the authors found significant within-group improvements with both intravitreal and subretinal rtPA and no clear difference between routes, but route comparisons remain vulnerable to confounding and treatment selection.

The multicenter STAR randomized controlled trial enrolled 90 eyes with nAMD-related SMH greater than two disc diameters, treated within 14 days. Participants received either PPV with subretinal rtPA and SF6 or pneumatic displacement with intravitreal rtPA; both arms received anti-VEGF. Mean BCVA gains were 16.8 versus 16.4 letters at three months (P = .767) and 17.2 versus 15.4 letters at six months (P = .776), respectively. Overall ocular adverse events were also similar in this trial (26.7% versus 27.3%).

This trial does not demonstrate visual superiority of PPV over PD in the enrolled population. It does not establish equivalence for every hemorrhage size, chronicity, or etiology, nor does it compare displacement against anti-VEGF monotherapy.

PPV with subretinal rtPA, gas tamponade, and ongoing anti-VEGF may be considered for extensive, thick, organized, or refractory SMH, or when less-invasive displacement is insufficient. In a 56-eye retrospective comparison, PPV with a subretinal therapeutic cocktail achieved foveal displacement in 87% versus 37.5% with PD-based treatment (P < .001), yet 12-month BCVA favored the PD cohort. This contrast illustrates why anatomic displacement and functional recovery must be assessed separately.

The Manchester protocol described in the review begins with intravitreal rtPA and gas, then escalates to PPV and subretinal rtPA if displacement is inadequate after approximately 72 hours. It is an observationally described stepwise strategy, not a universally validated standard of care.

Displacement procedures and surgery carry risks including rhegmatogenous retinal detachment (RRD), proliferative vitreoretinopathy (PVR), vitreous hemorrhage, RPE tears, and possible need for reoperation. A retrospective series of 167 eyes undergoing PPV with subretinal rtPA reported postoperative RRD in 15 eyes (9%); among eyes with RRD, 60% developed PVR. This is a selected surgical cohort and does not define the risk for every PPV technique or patient.

In a separate long-term 37-eye series using intravitreal rtPA, SF6, and bevacizumab, five eyes (13.5%) developed RPE tears. Comparative safety estimates vary substantially across study designs and hemorrhage severity. Discuss the individual patient’s baseline retinal status, lens status, macular condition, hemorrhage extent, and surgical complexity rather than relying on a single pooled complication figure.

• Confirm the cause and morphology: determine nAMD versus PCV and characterize hemorrhage size, thickness, foveal involvement, symptom duration, visual acuity, and associated neovascular activity using multimodal imaging.
• Smaller or moderate SMH: prompt anti-VEGF monotherapy is a reasonable strategy in suitable patients, with close follow-up.
• Larger or thicker recent SMH: consider adding PD with or without rtPA, especially where rapid foveal blood displacement appears feasible.
• Extensive or refractory SMH: discuss PPV and subretinal rtPA, recognizing that more complete anatomical clearance may not translate into superior visual acuity and may increase procedural risk.
• Throughout treatment: maintain appropriate anti-VEGF disease control, reassess OCT and vision, and use shared decision-making. This is a synthesis of the review’s themes, not a prospectively validated treatment algorithm.

• Narrative synthesis: no prespecified systematic search or de novo quantitative pooling is reported; conclusions depend on the selection and interpretation of cited studies.
• Confounding by indication: patients receiving PPV often have more severe, thicker, or later-presenting hemorrhage than patients treated with injections alone.
• Inconsistent definitions: studies vary in disc-diameter or area thresholds, thickness measurements, time to treatment, definitions of displacement, and follow-up.
• Evidence boundaries: STAR compares two displacement strategies rather than surgery versus anti-VEGF alone; TAPAS tests adjunctive rtPA/gas in a different population and setting.
• Clinical heterogeneity: PCV and typical nAMD may behave differently; evidence for traumatic, myopic, or RAM-related SMH is outside this review’s main focus.
• Ongoing research: the article describes TIGER as an ongoing randomized comparison of PPV/subretinal rtPA/gas plus aflibercept versus aflibercept monotherapy. It reports no TIGER outcome data; no trial results should be inferred from this review.

Prompt individualized management is more defensible than a universal intervention threshold. Anti-VEGF is foundational; PD and rtPA offer plausible and sometimes trial-supported benefit in selected larger hemorrhages; PPV remains valuable for extensive or refractory presentations but has not consistently improved final vision over less-invasive displacement. Clinical decisions should integrate hemorrhage burden, chronicity, underlying pathology, visual potential, procedural risk, and patient preference.